LinkedIn MCP servers
for Claude, compared.

Search linkedin mcp server and the top result is an open-source GitHub repo with roughly 3,700 stars: stickerdaniel's server, the one most developers wire into Claude first. It works, and it is comprehensive: the moment you connect it, every one of its tools, from a single profile lookup to a connection request, runs through your own logged-in LinkedIn session. Ask Claude to find two hundred heads of RevOps and start connecting, and it will go do exactly that, unattended, the moment you say so. That is the detail the search results bury: the real question is not just what these servers can do, it is whether they actually run the finding and the drafting for you, or just hand you an action and leave the rest to you.

So this is the developer's version of the answer. First, what a LinkedIn MCP server is and why Claude specifically makes it useful. Then the piece the SERP skips: a table of the servers that exist and what each one actually does. Then a worked transcript of running real prospecting, find to draft to approve, inside Claude, on a server built to carry the whole loop rather than one action at a time.

What a LinkedIn MCP server actually is

MCP, the Model Context Protocol, is the open protocol Anthropic introduced in late 2024 to let an AI model call external tools through a standard interface. An MCP server exposes a set of typed tools, an MCP client like Claude discovers them, and the model can then call them mid-conversation. Think of it as a shared socket between the model and whatever the server can do.

A LinkedIn MCP server, then, is any MCP server whose tools reach LinkedIn profiles, posts or actions. Ask Claude "find me heads of RevOps at B2B SaaS companies with 50 to 200 people" and, with a LinkedIn MCP server connected, the model can actually go do it and hand back structured results instead of guessing. The appeal is obvious. Unaided, Claude stops right at that action step, which is the whole subject of what Claude can and cannot do for sales: it researches and drafts well, then has nowhere to press send. The catch is that "reach LinkedIn" hides an enormous range of how, and the how is exactly what decides what you have to hand the server before it will do anything.

The LinkedIn MCP servers that exist, and what each one needs from you

Here is the single most useful thing to know before you install any of them. The servers on the SERP split by two things: what you have to hand each one to get it running, and how much of the actual prospecting job, finding and drafting, it does for you rather than leaving to you. This is the table to read twice.

The open-source repos are the ones most people find first, because they are on GitHub and they are free. They are also the ones that ask for your LinkedIn login up front. The most popular of them, stickerdaniel's server, is explicit about it: it either opens a browser for you to log in, imports cookies from your existing Chrome or Brave or Edge session, or saves a persistent authenticated profile, and then drives that session across its full toolset (server README on GitHub, as of 2026). Profile lookups, people search, messaging, connection requests, the feed, all of it runs as you.

That is a legitimate design and it works, and it is comprehensive rather than narrow: search, messaging, and connection requests all run through the one toolset. What it does not do is draft anything for you; that part is still on you or whatever you script around it.

The two servers that advertise a "LinkedIn API," felipfr's and quinnjr's, are worth a second look for a different reason. There is no public LinkedIn API that lets an outside developer search arbitrary people or message strangers, so a server offering those through "the LinkedIn API" is either using partner scopes it cannot actually grant you or reaching an undocumented endpoint behind the label. Read the auth step before you trust the noun: if it ends up asking for your credentials or a login token, that server is authenticating as you, API or not.

Put the idea to work

Less searching.
More conversations.

Tell Vera who you want to reach. Review the first message. Take it from there.

Ask Vera to find leads

Why capability matters more than the tool count

Why does capability matter more than the tool count? Because a long tool list can just mean many thin wrappers around the same one action, search, message, connect, with no layer that writes the first message for you.

There is no official third-party API that reads arbitrary profiles, runs people search, or messages strangers. We wrote the full map of what LinkedIn's partner API does and does not expose in the piece on LinkedIn API endpoints without approval, and the short version is that every one of those capabilities is off-limits to outside developers. So any LinkedIn MCP server that offers them is reaching them some other way, and there are only two other ways: work from public data, or replay a logged-in session.

A server built the second way has one job: replay whatever action you tell it to, as you. It is often comprehensive at that job, but the searching and the writing of a first message stay yours to do, either by hand or in whatever you script around the calls. Agent runtimes meet the same split: our piece on what an OpenClaw agent can reach on LinkedIn maps which routes need your signed-in browser.

Running LinkedIn prospecting inside Claude with BeReach

BeReach is built to run the whole loop rather than one action at a time: the find and draft workflow is done for you rather than assembled from individual tool calls you have to orchestrate yourself.

The connector is a hosted MCP server at mcp.bereach.ai that runs inside Claude.ai and Claude Cowork over Streamable HTTP, with OAuth 2.1: you sign in through your browser, so there is no key to paste into a config file. Getting it running is four steps:

1. In Claude, add a custom connector pointing at mcp.bereach.ai/mcp.
2. Claude opens the OAuth consent screen; sign in to BeReach and approve.
3. The tools load: ask it to find people and draft a first message for each.
4. When you want to send, connect your LinkedIn in Claude through Chrome. It is optional, only for invitations, messages and your own profile.

How to add it is in the docs: Claude and the API. The same address works in ChatGPT and other MCP clients. Under the connector there are 84 API operations behind a BEREACH_API_KEY if you would rather call the REST layer directly.

The 32-tool surface, grouped by what it touches

The connector's tools sort by job: most work on public data, and the outreach ones use your own LinkedIn, connected in Claude through Chrome.

Finding, reading and drafting work on public data. Sending, the inbox and connections are optional and use your own LinkedIn, connected in Claude through Chrome. The full list, what each tool does and whether it needs your account, is in the docs.

A worked transcript: find, draft, approve, send

Here is what the four-step loop looks like as an actual Claude conversation. The prompts are yours; the bracketed notes describe what BeReach does when Claude calls the tool.

You: Find founders and heads of growth at seed-to-Series-A B2B SaaS companies who posted about hiring their first SDR in the last month.

Claude: [calls find posts and find people on public data] I found 27 people matching that. They are founders and growth leads whose recent public posts mention hiring a first sales rep. Want me to draft a first message for each?

You: Draft a first-touch connection note for each of the 27. Reference the hiring post, keep it under 300 characters, no pitch in the note itself.

Claude: [calls draft dm for each contact] Drafted 27 notes. Each opens on the person's own hiring post and asks a question rather than pitching. They are sitting in drafts for your review; nothing has been sent.

You: Show me the first five, then I will approve.

Claude: [returns the drafts] Here are five. Edit any of them inline, or tell me to approve the batch.

You: Approve all 27.

Claude: [calls the send tool] To send, connect your LinkedIn in Claude through Chrome. Once it is connected, I will send the 27.

Notice what happened before anything went out: Claude found twenty-seven people, drafted a note for each, and you reviewed five before approving the batch.

If you want the same loop outside a live chat, that is the subject of the guide to an AI agent for LinkedIn outreach.

Which LinkedIn MCP server should you use

Match the server to the job, and the choice falls out of the table above.

• You want to read and manage your own LinkedIn from Claude, running the searches and actions yourself: an open-source browser-session server like stickerdaniel's does that, for free.
• You want raw profile or company data at scale and you do not need a LinkedIn account for it: a hosted scraper like Apify's returns data for an Apify token, no LinkedIn login, though you are scraping and paying per run.
• You want to run actual prospecting end to end, find, draft and approve: that is the case BeReach is built for. It builds the list, drafts a first message for each person, and a human approves every one before it goes out. It is the one paid option here rather than a free repo or a per-run scraper, so the trade is a subscription for a maintained public-data lane and a first message drafted for each person; current plans are on the pricing page.

You can also just try the find step with nothing installed. The describe the buyer, get the searches runs on the same public data the connector uses, in your browser, with nothing to connect.

In short, a LinkedIn MCP server lets Claude call LinkedIn tools mid-conversation. The ones you find first are open-source repos you host and configure yourself. The hosted options split further: some return raw data and leave the workflow to you, official connectors expose only LinkedIn's narrow sanctioned actions, and a connector like BeReach carries the whole find-and-draft loop end to end. Several outreach tools ship an MCP of their own: HeyReach competitors compared shows what each one does in Claude, and LinkedIn outreach tools compared covers the rest.

Frequently asked questions